Reassigning to arthursonzogni@, as I won't have time to look into these failures soon.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d100bb101d90652a502b50edc67b4676c870cce4

commit d100bb101d90652a502b50edc67b4676c870cce4
Author: Lukasz Anforowicz <lukasza@chromium.org>
Date: Thu Sep 28 02:43:04 2017

Exclusions for PlzNavigate issues found on the Site Isolation Win bot.

NOTRY=true

Bug:  769502 
Bug:  769508 
Change-Id: Ib894feba443f5bbfbcd133f0ad73b9dd12e962ab
Tbr: alexmos@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/688223
Reviewed-by: Alex Moshchuk <alexmos@chromium.org>
Commit-Queue: Łukasz Anforowicz <lukasza@chromium.org>
Cr-Commit-Position: refs/heads/master@{#504875}
[modify] https://crrev.com/d100bb101d90652a502b50edc67b4676c870cce4/third_party/WebKit/LayoutTests/FlagExpectations/site-per-process

I can repro. I am investigating....

This layout test is equivalent to:
----------------------------------------
<script>
  testRunner.waitUntilDone();
  testRunner.dumpAsText();
  window.location.href="http://127.0.0.1:1234/";
  setTimeout(()=>testRunner.notifyDone(),0);
</script>
----------------------------------------

With OOPIF, the current RenderFrameImpl is swapped out when it is asked to navigate to "http://127.0.0.1:1234".

1) testRunner.notifyDone() causes ShellViewHostMsg_TestFinished to be sent, but it will be canceled here:
----------------------------------------
bool RenderWidget::Send(IPC::Message* message) {
  // Don't send any messages after the browser has told us to close, and filter
  // most outgoing messages while swapped out.
  if ((is_swapped_out_ &&
       !SwappedOutMessages::CanSendWhileSwappedOut(message)) ||
      closing_) {
    delete message;
    return false;
  }
  [...]
----------------------------------------

In parallel with 1), I think the second RenderFrame stops loading and then the BrowserMainRunner stops.

According to me, there is a race condition in the test. We can't say whether the navigation will happens before or after testRunner.notifyDone().

If OOPIF is enabled, we can execute testRunner.notifyDone() in the first process and the navigation in the second.
With PlzNavigate, the navigation goes in the browser process faster.
That may explain why it behave differently.

What do you think? I think we should update the test.
I wonder if this test is up to date. (created: 7 years ago, last-modification: 5 years ago).
Can we request a navigation in the OnBeforeUnload?

Updating the test SGTM.  I don't know about requesting a navigation in OnBeforeUnload, but maybe other UI>Browser>Navigation folks can chime in.

Nasko found that this test is obsolete and is removing it in https://chromium-review.googlesource.com/c/chromium/src/+/762520/, so reassigning to him to close this bug once that CL lands.

The test is now gone.